A chaos-based model for low complexity predictive coding scheme for compression and transmission of electroencephalogram data.

A method for low complexity, low bit rate transmission of EEG (electroencephalogram) data, based on chaotic principles, is presented. The EEG data is assumed to be generated by a non-linear dynamical system of E dimensions. The E dynamical variables are reconstructed from the one-dimensional time series by the process of time-delay embedding. A model of the form X[n + 1] = F(X[n], X[n - 1], ... , X[n - p]) is fitted for the data in the E-dimensional space and this model is used as predictor in the predictive coding scheme for transmission. This model is able to give a reduction of nearly 50% of the dynamic range of the error signal to be transmitted, with a reduced complexity, when compared to the conventionally used linear prediction method. This implies that a reduced bit rate of transmission with a reduced complexity can be obtained. The effects of variation of model parameters on the complexity and bit rate are discussed.